We investigated the thermal effects caused by mobile phones during phone conservations (i.e temperature change during the conduction) on cortex. Please be aware that: - The issues regarding SAR Values is beyond the scope. - ‘Near Field Radiation Pattern’ and variaties about ‘Electric Fields’ is not discussed. - Mobile Phone, for this study, is considered to be a whole physical medium with certain properties (e.g insulator effect) rather than a device merely emitting RF signals
For this research both in-Vivo and in-Vitro models are used. We manufactured a phantom for in-Vitro modeling that can mimic the human tissue (i.e epidermal properties) The phantom produced using a specific recipe contains following ingredients:
During the phone call, following procedure is followed:
Having considered approximate body temperature is around 36.5 Celcius, we expected somewhat close values for the zeroth minute temperature value. Yet, it should also be noted that, 36.5 (or similar values) could be different depending on sex, age, and environmental conditions. Since face region is almost always affected by the outer world, certain low-level drops at the value could be just possible. On the other hand, for in-vitro process, we were able to keep track of the zeroth minute value to keep things under control. Gelatine based phantoms are likely to corrupt at high temperatures due to the nature of the preservative ingredient we used at manufacturing process (i.e alcohol). Therefore, initial temperatures for in-Vitro model were kept quite stable around 15 Celcius.
We generated a dataset composed of following columns: - Model Type - Temperature Values for each 5th minutes (e.g 5-10-15) including a initial value (i.e zeroth minute)
R (and Rstudio for IDE) is the language of choice for this project. Dataset is provided for reproducibility.
experiment <- data.frame(Model = c("H_1", "H_2", "H_3", "H_4", "H_5", "H_6", "H_7", "H_8", "H_9",
"H_10", "H_11", "H_12", "H_13", "H_14", "H_15", "H_16",
"H_17", "H_18", "H_19", "H_20", "P_1", "P_2", "P_3", "P_4", "P_5",
"P_6", "P_7", "P_8", "P_9", "P_10", "P_11", "P_12", "P_13", "P_14",
"P_15", "P_16", "P_17", "P_18", "P_19", "P_20"),
init = c(35.1, 35.6, 35, 35.4, 34.9, 35.2, 35.5, 34.7, 34.8, 34.9, 35,
34.9, 34.6, 35, 35.2, 35.5, 34.7, 34.9, 35.3, 35.2, 15, 15.2,
15.3, 15.2, 15, 15.4, 15.2, 15.3, 15.2, 15.1, 15.2, 15.1, 15,
15, 15.5, 15.2, 15.4, 15.2, 15.3, 15.2),
min5 = c(35.3, 35.6, 35.2, 35.5, 35.3, 35.6, 35.8, 34.9, 35, 35, 35,
34.7, 34.5, 35.3, 35.5, 35.5, 35, 35.1, 35.5, 35, 15.2, 15.2,
15.5, 15.6, 15.3, 15.5, 15.5, 15.5, 15.2, 15.4, 15.5, 15.2,
15.4, 15, 15.7, 15.5, 15.5, 15.3, 15.5, 15.8),
min10 = c(35.7, 35.9, 35.5, 35.8, 35.8, 35.5, 36, 35, 35.2, 35.4, 35.3,
35, 34.9, 35.7, 35.6, 35.5, 35.5, 35.4, 35.8, 35.4, 15.4,
15.2, 15.5, 15.9, 15.4, 15.8, 15.5, 15.9, 15.4, 15.5, 15.7, 15.6,
15.5, 15.2, 15.9, 15.7, 15.5, 15.3, 15.8, 15.9),
min15 = c(36, 36.2, 35.6, 35.7, 36, 35.9, 36.4, 35, 35.3, 35.5, 35.7,
35.3, 35, 36, 35.9, 35.8, 35.8, 35.5, 35.8, 35.5, 15.5, 15.7,
15.8, 16, 15.7, 16, 15.8, 15.9, 15.4, 15.5, 15.7, 15.6, 15.8,
15.6, 16, 16, 15.9, 15.7, 15.9, 16.1),
min20 = c(36.2, 36.3, 35.8, 36, 36, 36.1, 36.4, 35.5, 35.7, 35.8, 35.7,
35.5, 35, 36.4, 36, 35.8, 35.5, 35.9, 35.9, 35.8, 15.5, 15.7,
16, 16.4, 16, 16.3, 16, 16, 15.7, 15.7, 15.8, 15.6, 16, 15.7,
16, 16, 16.1, 16, 16, 16.3),
min25 = c(36.7, 36.3, 35.7, 36, 36.3, 36.1, 36.7, 35.7, 35.6, 36, 35.9,
35.9, 35.3, 36.5, 36, 36.2, 35.9, 36, 35.8, 35.8, 15.7, 16,
16.4, 16.5, 16.1, 16.4, 16.1, 16.3, 15.7, 16, 16, 15.7, 16, 15.6,
16, 16.3, 16.4, 16.3, 16, 16.4),
min30 = c(36.3, 36, 35.5, 36, 36.2, 35.9, 36.8, 35.5, 35.6, 35.7, 36,
35.7, 35, 36.5, 36, 36, 35.9, 36, 35.9, 36, 16.2, 16.5, 16.7,
16.8, 16.5, 16.7, 16.4, 16.5, 15.9, 16.2, 16, 15.9, 16, 15.8,
16.2, 16.1, 16.5, 16.5, 16, 16.5)
)
Load up the required packages
## ── Attaching packages ──────────────────────────────────────────────────── tidyverse 1.2.1 ──
## ✔ ggplot2 3.1.0 ✔ purrr 0.3.2
## ✔ tibble 2.1.1 ✔ dplyr 0.8.0.1
## ✔ tidyr 0.8.3 ✔ stringr 1.4.0
## ✔ readr 1.3.1 ✔ forcats 0.4.0
## ── Conflicts ─────────────────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
At this point, we need to wrangle our data to make it appropriate for the visualisation purposes by following the ‘tidy data’ procedure and pre-processing it.
Setting up the data types: it is another essential process to make an accurate analysis and to avoid obscure notation for possible future audits.
Now, since data is transformed, we explored the recently pre-processed dataset by choosing random samples out of it. Afterwards, we derived a grouped data by aggregation which might be quite useful for communicating the data.
## Observations: 280
## Variables: 4
## $ ModelType <fct> Human, Human, Human, Human, Human, Human, Human, Hum…
## $ ProcessNo <fct> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 1…
## $ Minute <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ Temperature <dbl> 35.1, 35.6, 35.0, 35.4, 34.9, 35.2, 35.5, 34.7, 34.8…
## ModelType ProcessNo Minute Temperature
## 1 Phantom 13 15 15.8
## 2 Phantom 14 5 15.0
## 3 Phantom 3 15 15.8
## 4 Phantom 6 15 16.0
## 5 Phantom 4 25 16.5
## 6 Phantom 8 25 16.3
## 7 Human 3 15 35.6
## 8 Phantom 10 20 15.7
## 9 Phantom 1 30 16.2
## 10 Phantom 4 10 15.9
## 11 Human 12 25 35.9
## 12 Phantom 6 30 16.7
## 13 Human 18 0 34.9
## 14 Phantom 11 30 16.0
## 15 Phantom 16 15 16.0
## 16 Human 4 30 36.0
## 17 Phantom 10 25 16.0
## 18 Human 15 20 36.0
## 19 Phantom 19 0 15.3
## 20 Phantom 14 20 15.7
## # A tibble: 8 x 4
## # Groups: ModelType [2]
## ModelType Minute meanTemp sdTemp
## <fct> <dbl> <dbl> <dbl>
## 1 Human 10 35.5 0.31
## 2 Human 0 35.1 0.290
## 3 Human 15 35.7 0.37
## 4 Human 30 35.9 0.38
## 5 Phantom 5 15.4 0.19
## 6 Phantom 30 16.3 0.3
## 7 Phantom 10 15.6 0.23
## 8 Phantom 25 16.1 0.27
featuring ggplot2 and lattice packages We created insightful visualisations to make interpretations on the findings and possible reasons.
Similar graphs with a more recent and modern package - ggplot2:
A more simplified graph was produced via statistical averages of temperature values
Here, we can have some interactivity for demonstration considerations.
We will get back to the data and filter it to work with in-Vivo values only, due to the fact that main topic is a lot more relevant to human beings. We are going to generate some new variables to make further analysis and derive some informations from them in order to figure out what the temperature change is on the surface of the brain.
Please note that this type of effect can only be found numerically by performing quite complicated analysis and solving very complex analytical equations! (e.g bioheat transfer equations)
Our study DOES NOT contain that type of process. Instead, we revised well-respected papers addressing this issue and derive a numeric COEFFICIENT to make some estimations. Therefore, the end-results you are about to evaluate are just rough approximations. NOT an end-to-end analysis!
## # A tibble: 5 x 5
## ProcessNo maxDegree minDegree maxDelta expHighChange
## <fct> <dbl> <dbl> <dbl> <dbl>
## 1 18 36 34.9 1.1 0.094
## 2 1 36.7 35.1 1.6 0.137
## 3 6 36.1 35.2 0.900 0.077
## 4 9 35.7 34.8 0.9 0.077
## 5 7 36.8 35.5 1.30 0.111
So, which process generated maximum (or minimum) temperature change?
Finally, despite of the fact that around 3.5 Celcius degree is considered critical by medical experts, studies have shown that hipotalamus, in charge of thermal regulations, is likely to alter the thermal behaviour of body after exceeding a change of 0.2 celcius degree. Let’s see if one (or more) of our processes have reached that threshold.
For in-vitro models, there is nothing shocking especially considering the non-alive nature of the material. We have observed an almost linear rise for the phantom which is quite expected. We also expect that after some point the line is very likely to lose its trend and level-off presumably due to molecular resistance. For in-vivo process, on the other hand, a quite interesting trend can be observed. Temperature somewhat aggressively rises around 20-25th minute and then levels-off and starts a smooth declining trend.
Despite being no observation around the aforementioned threshold of 0.2 Celcius Degree at our final figure, we have seen a potential approach to that point. Therefore, there is no harm in taking simple pre-cautions to avoid possible issues.